Comparison of live birth defects after luteal phase ovarian stimulation vs conventional ovarian stimulation for in vitro fertilization and vitrified embryo transfer cycles

No evidence was found of detrimental effects of luteal-phase ovarian stimulation on live-born infants at birth. Infertility itself and multiple births pose risk factors for congenital malformation.

Hong Chen, M.D., Yun Wang, M.D., Qifeng Lyu, Ph.D., Ai Ai, M.D., Yonglun Fu, M.D., Hui Tian, M.D., Renfei Cai, M.D., Qingqing Hong, M.D., Qiuju Chen, Ph.D., Zeev Shoham, M.D., Yanping Kuang, M.D.

Volume 103, Issue 5, Pages 1194-1201


To assess live-birth defects after a luteal-phase ovarian-stimulation regimen (LPS) for in vitro fertilization (IVF) and vitrified embryo transfer (ET) cycles.

Retrospective cohort study.

Tertiary-care academic medical center.

Infants who were born between January 1, 2013 and May 1, 2014 from IVF with intracytoplasmic sperm injection (ICSI) treatments (n = 2,060) after either LPS (n = 587), the standard gonadotropin-releasing hormone–agonist (GnRH-a) short protocol (n = 1,257), or mild ovarian stimulation (n = 216).

The three ovarian-stimulation protocols described and assisted reproductive technology (ART) treatment (IVF or ICSI, and vitrified ET) in ordinary practice.

Main Outcome Measure(s):
The main measures were: gestational age, birth weight and length, multiple delivery, early neonatal mortality, and birth defects. Associations were assessed using logistic regression by adjusting for confounding factors.

The final sample included 2,060 live-born infants, corresponding to 1,622 frozen–thawed (FET) cycles, which led to: 587 live-born infants from LPS (458 FET cycles); 1,257 live-born infants from the short protocol (984 FET cycles); and 216 live-born infants from mild ovarian stimulation (180 FET cycles). Birth characteristics regarding gestational age, birth weight and length, multiple delivery, and early neonatal death were comparable in all groups. The incidence of live-birth defects among the LPS group (1.02%) and the short GnRH-a protocol group (0.64%) was slightly higher than in the mild ovarian-stimulation group (0.46%). However, none of these differences reached statistical significance. For congenital malformations, the risk significantly increased for the infertility-duration factor and multiple births; the adjusted odds ratios were 1.161 (95% confidence interval [CI]: 1.009–1.335) and 3.899 (95% CI: 1.179–12.896), respectively. No associations were found between congenital birth defects and various ovarian-stimulation regimens, maternal age, body mass index, parity, insemination method, or infant gender.

To date, the data do not indicate an elevated rate of abnormality at birth after LPS, but further study with larger populations is needed to confirm these results. However, infertility itself poses a risk factor for congenital malformation. A higher likelihood of birth defects in multiple births may lead couples to favor elective, single ET; couples undertaking ART should be made aware of the known increased birth defects associated with a twin birth.

  • Shvetha Zarek

    As the authors highlight, there is a low frequency of birth defects (n=15 out of 2,060 live born infants) in the overall cohort, especially when sub-divided by the three stimulation methods of interest. Thus, there is a strong possibility that this cohort is underpowered to detect differences between the three methods, suggesting a possible type II error. In addition, the majority of the birth defects occurred in multiple gestations (10/15). Thus, the only supported conclusion, (that is already well established) is that multiple gestation increases the risk of birth defects. A conclusion that LPS does not indicate an elevated rate of birth defects is premature and cannot be made from this data. The authors correctly highlight that larger studies are needed. The authors should kindly consider excluding multiple gestation pregnancies in future studies. I welcome any comments from the authors. Thank you.

Translate »